Printed on the page
Loan applications, Loan Estimates, Closing Disclosures
Read straight from the document, like any other field.
CapturePoint 6 for banks, credit unions and lenders
CapturePoint 6 finds where each document starts, knows a Loan Estimate from a Closing Disclosure from a pay stub, reads the values that matter and names every file. It runs on the branch PC, and sets itself up from a folder of your own loan files.
Free trial starts the first time you open it. No form.
Try it in five steps
Windows 10 or 11, 64-bit. The installer is about 367 MB, with a one-time download of about 6 GB the first time it runs. The free trial starts on first launch, with no form.
On the welcome screen choose Take a 2-minute tour. You do the real steps on five sample accounts payable documents.
Choose Get started and bring in a folder of loan files you already have. It sets up the document types and fields from them.
Most documents go straight through. The few that wait say why, in plain words. Confirm one with Ctrl+Enter.
Send the documents to a folder and open them: one searchable PDF per document, named from the values it read.
What it reads
Set up from a folder of your own loan files, it reads values such as these. Anything else in your files, from a flood determination to your own checklist, it learns from the ones you confirm.
The question every loan processor asks
Loan applications, Loan Estimates, Closing Disclosures
Read straight from the document, like any other field.
Any package you scan as a unit
Scan each package behind a barcode cover sheet that carries the loan number. Set the job to start documents automatically and at matching barcodes.
Pay stubs, W-2s, bank statements, VOEs
Import an export of active loans. The borrower name read from the page finds the loan number in the list. A name that fits two loans waits for a person.
What waits for a person
You decide, field by field, how sure it must be before a value goes without review. Before you save, it shows how many of your recent documents would go straight through and how many would wait.
On the review screen the page, its values and any table sit side by side, with the reason it waited. Correct what is wrong and press Ctrl+Enter to confirm the whole document.
Every confirmation teaches it. After you confirm a few, it checks the documents still waiting again, and the ones that now pass clear by themselves. After steady use, most loan packages need no review at all.
Two pay stubs in a row should be two documents. If it ever joins two documents or splits one, you correct where the document starts in review, and that teaches it too.
Named and filed
Each document type gets its own folders and file name, built from parts: the document type, any value it read, or your own words. This example uses one folder per loan number, then the document type and the person's name: the borrower on an application or disclosure, the employee on a pay stub.
Two files that would share a name get -2, -3, so nothing is overwritten. For each document you can have a searchable PDF (PDF/A if you archive), a data file with every value, and a text file.
Example output, fictional loan
Or send documents with their values to Content Central, where they join the loan file. Each one can also go to SharePoint or OneDrive (work or school accounts), Google Drive, Dropbox or Nucleus One.
The lender scans the package before the borrower leaves, so a missing pay stub is asked for in person. One station per branch, with unlimited scanning on each. Every branch can run the same job, and the same naming.
Branches send paper by courier and one faster scanner does the volume. Cover sheets printed at the branch keep each package together on the way.
Any TWAIN scanner on a Windows 10 or 11 PC. It also reads files you already have: PDF, TIFF, JPEG, PNG, BMP and GIF, so loan files scanned years ago can be split and named too.
No. CapturePoint 6 reads every page on the PC it is installed on. A graphics card makes it faster but is optional. Finished files go only where you send them.
Yes. Confirm or correct a few and it learns them, or point automatic job setup at a folder of samples and it works out the document types and fields itself. Every confirmation teaches it.
Each document can be exported with a data file holding every value it read, next to the searchable PDF, for another system to pick up from the folder. CapturePoint 6 also sends documents and values straight to Content Central, which can keep the whole loan file and show what is missing.
Then scan one real package of your own. Would rather see it with us first? We will run it live on a call.